About Me

Selly is now an software engineer at Google. Prior that, she had the experience in developing machine learning model integrated automation system for 2 years. She received her B.S. and M.S. degree both in Electrical Engineering under supervision of Prof. Chi-Chun(Jeremy) Lee from National Tsing-Hua University, Taiwan in 2017 and 2020. Her study emphasized on learning speech front-end network for emotion recognition.

Technical Strengths

Computer Languages Python, C/C++, MATLAB, Java, Javascript, PHP, Shell script
Software & Tools TensorFlow, Scikit-learn, PyTorch, Kaldi, LaTex, HTML, CSS

Work Experience

Google Taiwan Engineering Limited

Software Engineer in Silicon • Mar 2022 – Current

Qualcomm Semiconductor Limited

Machine Learning Engineer in Center for Operations, Manufacturing Engineering, and Testing (COMET) • Jul 2020 – Mar 2022

  • Build a wafer defect recognition model above 90% accuracy and integrate it with the automation system to recommend disposition.
  • Provide new automation solution for yield management team to recognize potentially failed dies which saves their 50% work time.
  • Mentor for 3 junior employees and interns in independently maintaining existing projects in three months.


National Tsing Hua University

M.S. in Electrical Engineering • Sep. 2017 – Apr. 2020

  • Thesis: "Observe Critical Data in Emotion Recognition Using a Speech Front-end Network Learned from Media Data In-the-Wild", supervised by Prof. Chi-Chun (Jeremy) Lee
  • Teaching Assistant: Music Information Retrieval (2019 Spring CS5731 by Dr. Li Su)
  • Honor Society: Phi Tau Phi(Top 3% of master’s graduands that are exellent in academic performance.)

National Tsing Hua University

B.S in Electrical Engineering • Sep 2013 – Jun 2017

  • Emphasis in digital signal processing
  • Research Assistant supervised by Prof. Chi-Chun Lee(May 2016 – Jun. 2017)


AI for Leukemia Diagnosis (AHEAD)

Machine Learning Scientist • May 2017 – Sep 2019

  • The recidual leukemia cell detection is 100+ times faster than traditional method and achieve 90% accuracy.
  • AML disease assessment and outcome prediction achieves around 80% accuracy for mortality and relapse prediction in next-3-month.
  • Exhibited at FutureTech2018, CES2019


May 2016 – Feb 2017

  • We built an FPS prediction model by analyzing over ten million gaming-hour data; and based on it, providing recommendation of PC elements to achieve the desired FPS.


  1. Chih-Chuan Lu, Jeng-Lin Li, Yu-Fen Wang, Bor-Sheng Ko, Jih-Luh Tang and Chi-Chun Lee, "A BLSTM with Attention Network for Predicting Acute Myeloid Leukemia Patient's Prognosis using Comprehensive Clinical Parameters", in Proceedings of Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2019 [Oral presentation] doi
  2. Bor-Sheng Ko, Yu-Fen Wang, Chih-Chuan Lu, Jeng-Lin Li, Chi-Chun Lee, Jih-Luh Tang, and Hwei-Fang Tien, "Relapse and Mortality Prediction of Acute Myeloid Leukemia Patients Using Deep Bidirectional Long Short-Term Memory-Deep Neural Network Architecture", in Annual Meeting of American Society of Hematology (ASH), 2018 doi
  3. Chih-Chuan Lu, Jeng-Lin Li, Chi-Chun Lee, "Learning an Arousal-Valence Speech Front-End Network using Media Data In-the-Wild for Emotion Recognition", in Proceedings of ACM on the Audio/Visual Emotion Challenge and Workshop (AVEC), 2018 [Oral presentation] doi